The proposed studies address temporal relationships of behavioral and neuropathological dysfunction in transgenic (Tg) mice expressing the amyloid precursor protein (APP) in different genetic settings. Different host strains influence the type and onset of behavioral abnormalities, the type and extent of neuropathology, and the longevity of Tg APP mice,, reflecting genetic variation between host strains that provides a rick source of information about genes that modify the phenotype of Tg mice expressing APP. However, these same strain related differences introduce experimental variability that could potentially thwart the reproducibility of scientific results. Identifying temporal windows of reversibility of phenotypic changes associated with expression of APP transgenes in mice may provide information about optimal timing of potential treatments. One interesting question about Alzheimer's disease (AD), particularly dominantly inherited AD, is why individuals at-risk remain neurologically well until many decades after birth. Two alternative hypotheses will be explored in Tg mice. First, we propose to identify an inbred strain of mouse in which to harbor a large APP transgene array. We will produce "speed congenics" using marker-assisted selection to produce the same level of congenicity at the fourth (N4) or fifth (N5) generation that normally takes 10 or 11 backcrosses. In an appropriate inbred strain, we will perform at closely spaced time intervals analysis of memory and learning, histopathology, and Abeta levels to asess the temporal inter-relationship of these traits. Second, we propose to test the hypothesis that amyloid deposits and/or deficits in learning and memory can be reversed by down-regulating expression of APP transgenes in adult mice. This shall be accomplished by developing a new type of Tg mouse harboring regulable APP transgenes using a tet-transactivator. Feeding mice doxycycline to suppress APP transgene expression at different times after the development of neuropathology of behavioral deficits will provide information about critical time windows allowing for reversibility of specific traits. Third, we propose to test the alternative hypothesis that delayed manifestation of disease traits in Tg2576 mice is caused either by chronicity of exposure or by an age-related biological change in the animal's ability to tolerate one or more toxic products involved with Tg APP expression. This shall be accomplished by developing Tg mice harboring regulable APP transgenes using a "reverse" tet-transactivator to study of the role of aging in AD. Feeding mice doxycycline to activate APP transgenes in mature or aging mice may allow us to distinguish between chronicity of exposure and a biological change in vulnerability related to aging. The observation that behavior and/or pathological traits ensure very shortly after activation in aging mice would favor the latter hypothesis.